Considerations to Optimize Structurally Responsive Work Environments for U.S. Military Specialty Physicians.

Abstract

Introduction: Recruitment, training, and retention of wartime critical specialty physicians may be stymied by discrimination and abuse. It is unclear to what extent the US combat specialty physicians witness or experience discrimination and abuse, whether they or others intervene, and if they would subsequently discourage people from entering their field.

Materials and methods: The present study surveyed US active duty anesthesiologists, emergency medicine physicians, and orthopedic surgeons (N = 243; 21% response rate). A generalized linear model evaluated the extent to which gender, specialty, service, and number of witnessed/experienced negative/stigmatizing comment/event types were associated with burnout. A bootstrapped mediation analysis evaluated whether gender and burnout were mediated by the number of comment/event types.

Results: The sample was majority non-Latine White (87%) and men (66%) with tri-service and specialty representation. The most commonly reported negative/stigmatizing comment/event types were related to pregnancy (62%) and parental leave (61%), followed by gender and assigned sex (42%), lactation (37%), and sexual harassment (35%). Of the respondents who witnessed/experienced such comments/events, self-intervention was reported after comments regarding lactation (43%), assigned sex and gender (42%), race and ethnicity (41%), pregnancy (41%), parental leave (37%), and sexual harassment (24%). Witnessing another person intervene was reported after sexual harassment (25%) and comments/events regarding race and ethnicity (24%), pregnancy (20%), assigned sex and gender (19%), lactation (19%), and parental leave (18%). Nonintervention was reported after comments/events related to parental leave (42%), pregnancy (38%), sexual harassment (26%), lactation (26%), assigned sex and gender (26%), and race and ethnicity (22%). Respondents reported moderate-to-high intervening likelihood, importance, and confidence. Respondents reporting neutral to extremely agree on prompts indicating that pregnant active duty physicians are trying to avoid deployment (P = .002) and expect special treatment that burdens the department (P = .007) were disproportionately men (36% and 38%, respectively) compared to women (14% and 18%, respectively). The highest proportion of neutral to extremely agree responses regarding discouraging specialty selection were reported in relation to transgender and gender diverse students (21%), followed by cisgender female students (18%); gay, lesbian, or bisexual+ students (17%); cisgender male students (13%); and racial and ethnic minoritized students (12%). In the primary model, the number of witnessed/experienced comment/event types was associated with greater burnout (0.13, 95% CI 0.06-0.20, P = .001), but women did not report significantly different levels of burnout than men (0.20, 95% CI -0.10 to 0.51, P = 0.20). The number of comment/event types mediated the relationship between gender and burnout (0.18, 95% CI 0.06, 0.34; P < .001).

Conclusions: Although reported intervening confidence, likelihood, and importance were high, it is unclear whether perceptions correspond to awareness of intervention need and behaviors. Bivariate differences in burnout levels between men and women were fully explained by the number of comment types in the mediation model. Annual trainings may not effectively address workplace climate optimization; institutions should consider targeted policy and programmatic efforts to ensure effective, structurally responsive approaches.